Modelagem do fenômeno de flutter via CFD: aplicação a estrutura da ponte Great Belt east

Abstract

Slender structures, such as large-span bridges, tend to have greater aeroelastic sensitivity, which directly influences the project from its conception. The flutter phenomenon must be carefully evaluated, as it can lead these structures to collapse. To obtain and analyze it, Computational Fluid Dynamics (CFD) can be used, which is constantly evolving and is increasingly used by researchers to obtain reliable results through numerical methods. In this study, a CFD methodology is presented to obtain the flutter coefficients, through forced vibrations in bodies immersed in a fluid medium. Two approaches are analyzed. In the first (8COEF), torsion and flexion movements are imposed on the structure. In the second, simulations are carried out in the torsion mode, and, through Linear Equations (LE), the coefficients of the flexion movement are obtained, thus resulting in the complete set of eight coefficients of flutter. The study is applied to the transverse section of the deck of the Great Belt East Bridge (GBEB), located in Denmark. For the static case, in a 2D approach, simulations were performed in order to obtain the average values of the aerodynamic coefficients of drag (𝐶𝑑), lift (𝐶𝑙) and moment (𝐶𝑚), in addition to the Strouhal Number (𝑆𝑡), adopting four Gauss divergence schemes: QUICK, upwind, linear upwind and limited linear. For the dynamic case, also in a 2D approach, simulations were performed to obtain the flutter coefficients. Then, the critical flutter velocity was estimated using the 8COEF and LE approaches. The simulations were carried out in a turbulent flow regime with 𝑅𝑒 105 , using the RANS approach (Reynolds Averaged Navier-Stokes), using the models of k-𝜔 SST and k-𝜔 SSTLM turbulence to establish a comparative analysis. OpenFOAM® was used to perform the computational simulations presented in this study and also to solve the Navier-Stokes Equations for an incompressible fluid. It is noted that the estimated results with the LE were effective when compared to the 8COEF but at a lower computational cost. Finally, all results were validated with numerical and experimental studies available in the literature.